Formation and Migration of Oxygen Vacancies in ... - ACS Publications

Jul 18, 2016 - King Abdullah University of Science and Technology (KAUST), Physical Science and Engineering Division (PSE), Thuwal 23955-6900, Saudi A...
0 downloads 0 Views 913KB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Formation and Migration of Oxygen Vacancies in SrCoO and their effect on Oxygen Evolution Reactions 3

Hassan A. Tahini, Xin Tan, Udo Schwingenschlögl, and Sean C. Smith ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b00937 • Publication Date (Web): 18 Jul 2016 Downloaded from http://pubs.acs.org on July 20, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

FORMATION AND MIGRATION OF OXYGEN VACANCIES IN SrCoO3 AND THEIR EFFECT ON OXYGEN EVOLUTION REACTIONS

Hassan A. Tahini1, Xin Tan1, Udo Schwingenschlögl2 and Sean C. Smith1* 1

Integrated Materials Design Centre (IMDC), School of Chemical Engineering, UNSW Australia, Sydney, NSW 2052, Australia

2

Physical Science & Engineering Division, KAUST, Thuwal 23955-6900, Kingdom of Saudi Arabia

ACS Paragon Plus Environment

1

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 18

ABSTRACT

Perovskite SrCoO3 is a potentially useful material for promoting the electrocatalytic oxygen evolution reaction, with high activities predicted theoretically and observed experimentally for closely related doped perovskite materials. However, complete stoichiometric oxidation is very difficult to realize experimentally – in almost all cases there are significant fractions of oxygen vacancies present. Here, using first principles calculations we study oxygen vacancies in perovskite SrCoO3 from thermodynamic, electronic and kinetic points of view. We find that an oxygen vacancy donates two electrons to neighboring Co sites in the form of localized charge. The formation energy of a single vacancy is very low and estimated to be 1.26 eV in the dilute limit. We find that a vacancy is quite mobile with a migration energy of ~0.5 eV. Moreover, we predict that oxygen vacancies exhibit a tendency towards clustering which is in accordance with the material’s ability to form a variety of oxygen-deficient structures. These vacancies have a profound effect on the material’s ability to facilitate OER, increasing the overpotential from ~0.3 V for the perfect material to ~0.7 for defective surfaces. A moderate compressive biaxial strain (2%) is predicted here to increase the surface oxygen vacancy formation energy by ca. 30%, thus reducing the concentration of surface vacancies and thereby preserving the OER activity of the material.

KEYWORDS: perovskite, electrocatalyst, oxygen evolution reaction, DFT, vacancies, strain

1. Introduction The search for materials with high efficiency to drive oxygen evolution reactions for electrochemical water splitting is intensely pursued for its obvious environmental benefits.1–3 The overall reaction is given by 2H O → O + 4H  + 4e and comprises a total four proton-electron transfer steps.4 To avoid any one step acting as a bottleneck for the overall reaction, each step might ideally require 1.23 V in order to

ACS Paragon Plus Environment

2

Page 3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

minimally match the thermodynamics of the overall process. This ideal is extremely difficult to achieve, however, and under operating conditions an applied overpotential is necessary for the reaction to proceed. To lower this overpotential, noble metals are used such as platinum and its alloys, which severely restricts the prospects of using these technologies on a commercial scale.2 Several alternatives to noble metals and their oxides have been put forward to catalyze OER.5,6 The perovskite family is a versatile family of compounds with a chemical formula ABO3. The A and B sites typically consist of alkali or alkali earth metals and transition metals, respectively, and each compound can host a range of atoms on the A or B sites acting as dopants which enriches its functional flexibility. Earlier work4,5,7 has established linear relationships between the binding energies of the various intermediates involved in the reaction, which imposes a restriction on the ability to lower the overpotential with a predicted lower bound of 0.2~0.4 V for oxides.5 Previous theoretical works have shown that in the family of the perovskites, SrCoO3 has theoretically the lowest overpotential of~0.25 V.5,8 However, this has been hard to demonstrate experimentally due to the fact that this material does not easily form in the ideal SrCoO3 pm3m structure but rather as an incompletely oxidized SrCoO3-δ, where δ=0.5 is the most common phase, adopting an orthorhombic crystal structure.9 Suntivich et al.10 demonstrated that Ba0.5Sr0.5Co0.8Fe0.2O3-δ (BSCF) can outperform the best know catalysts RuO2 or IrO2, which was explained using a descriptor based on the filling of the eg orbitals. However, BSCF undergoes a transformation at relatively low temperatures into new phases that do not exhibit the same performance as the parent compound.11 Recently, it was demonstrated that SrCoO3 could be stabilized in a cubic crystal form by adding scandium (Sc) and niobium (Nb) substitution to the B site.12 These additions where shown to enhance the electrocatalytic activity of SrCoO3 by about 6 times over BSCF at an overpotential of 0.35 V.

In this work, we study the oxygen vacancy formation and clustering in bulk SrCoO3 as well as defective surfaces in an attempt to understand the effects of vacancies on OER. We show that vacancies in the vicinity of surface Co sites can lower their catalytic activity. To mitigate the formation of

ACS Paragon Plus Environment

3

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 18

detrimental vacancies we propose the application of biaxial strain which suppresses VO formation while preserving the OER performance.

2. Methodology Our calculations are based on VASP’s implementation13 of density functional theory with the projector augmented waves method14 to treat core and valence electrons. The generalized gradient approximation with the PBEsol15 parameterization was used to describe the exchange-correlation, along with the RPBE16 functional which gives a better description of adsorption energies and was used previously in similar studies.4,5,7 The RPBE functional also reproduces O2 bond lengths (1.23 Å) and H2O bond lengths and angles (H-O=0.97 Å, ∠H-O-H=103.6o) that are in agreement with previous theoretical calculations performed at a different level of theory.17 Plane waves with kinetic energy cutoff of 500 eV were used to expand wavefunctions. Energies were converged with a 1x10-5 eV tolerance and forces were optimized until they change by no more than 1x10-2 eV/Å. Adsorbed species were modelled using a 6-layer 2x2 slab with a vacuum separation of about 15 Å (selected 4x4 surfaces where included for comparison in the supporting information). Dipole corrections were included in this study.18

3. Results and Discussions Defect free system- We begin by looking at the structural and electronic properties of the defect free SrCoO3. SrCoO3 with a nearly full stoichiometry was found to exhibit metallic behavior along with ferromagnetic ordering.19 The formal oxidation state of the B sites in an ABO3 compound is generally determined by the oxidation state of the A site, which in this case leaves Co with a 4+ charge, which allows for the possibility of several spin states due to the presence of unpaired electrons.20 Magnetic moments ranging from 1.5  to 3.0  have been predicted.20 In our model the calculated magnetic moment was found to be 1.8 μ . This low magnetic moment qualitatively agrees with the intermediate

ACS Paragon Plus Environment

4

Page 5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

spin (IS) point of view,21 but still it is underestimated due to the GGA functional used when compared with that obtained with hybrid functional for instance.20 Experimentally, SrCoO3-δ exists in the cubic phase for low oxygen deficiencies. Taguchi et al.19 determined the lattice parameter to be 3.836 Å. This value did not deviate from the those obtained by more modern and recent experiments.22,23 By fully relaxing the unit cell we obtained a cubic structure with a lattice parameter of 3.775 Å with PBEsol. On the other hand, PBE and RPBE predict a lattice parameter of 3.838 Å and 3.833 Å, respectively, which is in agreement with previous theoretical calcualtions.20 In this work we will employ the PBEsol functional to study the oxygen vacancies and their diffusion. These bulk results will be compared to those obtained using RPBE which will later be used to study the adsorption of OER intermediates.

Oxygen vacancies- We simulate the effect of oxygen vacancies in SrCoO3 using a 2x2x2 supercell by removing one O atom from its normal site. This creates an O-Co-VO oriented along the z-axis to simplify orbital analysis (by aligning the local d-orbitals with the global projection axis used by the DFT code). The removal of one neutral O atom leads to two electrons being left behind in the lattice. Generally, these electrons could end up localized on neighboring metal sites or delocalized through the lattice.

ACS Paragon Plus Environment

5

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 18

Figure 1. (a),(b) Differential charge density upon the creation of VO and a slice along the (100) plane. (c) The d3z2-r2 and dx2-y2 contributions of the Co ions in a pristine bulk system and the Co ions surrounding VO. Figures were rendered using VESTA.24

The charge density redistribution caused by the creation of the vacancy is given with respect to the charge density of the vacancy containing supercell [ρ(VO)], the bulk system [ρ(bulk)] and a single O atom added at the vacancy site [ρ(O)] and is given by ∆ρ = ρ(VO)+ρ(O)- ρ(bulk). The resulting differential charge density is shown in Fig.1(a),(b). Here, it is clear that VO donates charge to the two neighboring Co ions. The shape of the accumulated charge suggest that the donated electrons go into   orbitals centered on the Co sites, while the depleted charge regions on the O ion suggests that this charge originates from the py orbitals. Further evidence for charge accumulating in the   orbitals is given by the DOS plots shown in Fig.1(c). The local DOS (LDOS) for the Co ions closest to the vacancy show

ACS Paragon Plus Environment

6

Page 7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

a peak right below the Fermi level, for which the integrated profile between -0.5 and the Fermi level gives 1



.

The oxygen formation energy was calculated according to: 1 !" #$% ) = !#$% ) & !#perf) + )Δ+ + !#O ), 2 Where !#$% ), !#perf) and !#O ) are the total energies of the defect containing supercell, perfect supercell and the energy of an O2 molecule. Δ+ is a correction term added to account for the well-known O2 overbinding in DFT calculations. This value is obtained by fitting the heats of formation of a range of oxides and determining the offset from the well characterized experimentally determined heats of formation.25,26 Using this approach we find the correction with PBEsol to be 0.9 eV/O2 (1.7 eV/O2 using RPBE and 1.4 eV/O2 using PBE which is consistent with previous theoretical calculations25–27). The vacancy formation energy was computed for a number of supercell sizes which then allows the extrapolation to a dilute limit. These are, the 2x2x2, 3x3x3 and 4x4x4 cubic cells and the -√2 / -√2, 2-√2 / 2-√2 and the 3-√2 / 3-√2 rhombohedral cells. The obtained formation energies are then fitted to !" #0) = !1 + -0 2 + 30  to account for electrostatic and elastic interactions.28,29

ACS Paragon Plus Environment

7

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 18

Figure 2. The oxygen vacancy formation energy calculated as a function of cell size using cubic and rhombohedral cells. The data points are fit into a polynomial of the form !" #0) = !1 + -0 2 + 30  and is represented by the solid line.

The extrapolated value to infinity (Fig. 2) puts the formation energy of a single vacancy at 1.26 eV in the oxygen rich regime (1.17 eV with RPBE). This is much lower than other typical perovskites. For instance in LaNiO3 this value was determined to be 2.28 eV29 and 2.7 eV for La0.75Sr0.25MnO3.30 On the other hand, the value is more comparable to BSCF which was found to be 1.34~1.40 eV11 (depending on local environment). This low formation energy suggests a high concentration of VO even under O-rich conditions. VO concentration will further increase at lower partial pressures of oxygen signalling a preference to form non-stoichiometric SrCoO3-δ.

Oxygen divacancy interactions- To shed some light on the interaction of oxygen vacancies in SrCoO3, we used a 2x2x2 supercell consisting of eight unit cells. We considered two VO at a time. The atomic positions were relaxed without taking into account the distortions and deformations to the supercell volume and shape that might be caused by the presence of the vacancies, which we will consider separately. The chosen supercell offers seven unique arrangements of the VO (see Fig. 3).29,31 To assess the extent of the vacancy interactions, we calculated the binding according to: !4 #2$5 ) = !#2$5 ) & 2!#$5 ) + !#perf) Where !#2$5 ) is the total energy of the system containing a divacancy. Here, a negative binding energy corresponds to a stable cluster relative to two VO that are infinitely separated.

ACS Paragon Plus Environment

8

Page 9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 3. The left panel depicts the divacancy configurations labelled A-G. Shown on the right are the binding energies corresponding to each configuration. The shaded planes represent the positions of the oxygen vacancies.

Fig. 3 shows that several configurations are stable, in particular configuration C depicting two vacancies in the same plane in a basal site of the CoO6 octahedron. The next most stable cluster is configuration A which represents an apical divacancy (the two vacancies are separated by a Co ion). Previous studies have showed that this is the stable configuration in SrTiO331 and LaNiO329. The tendency of VO to exist in bound clusters is in agreement with the established observation that SrCoO3 can easily form oxygen deficient structures. Moreover, the preference for certain orientations such as the basal or apical divacancy configurations could explain the transition to brownmillerite phase which is a vacancy ordered structure.9,32,33

Oxygen migration in SrCoO3- In general, theoretical calculations have revealed that for a number of perovskites O ions migrate from one vacant site to another via a curved migration pathway.34,35 We have confirmed that this is also the case for SrCoO3 (see Fig. 4(b)). Our calculations were performed in a 40, 135 and 320 atom supercells and the migration energies were found to be 0.78 eV, 0.5 eV and 0.56 eV, respectively. These values are still lower than the oxygen migration in SrCoO2.5 (δ=0.5) in the

ACS Paragon Plus Environment

9

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 18

brownmillerite phase which has a different crystal structure but nevertheless a similar oxygen environment for the octahedral layers in which the calculated barrier is 0.82 eV.36 However, our values are similar to those in Ba0.5Sr0.5CoO3 (0.40 eV).37

Figure 4. (a) The migration energy barriers for VO in a 40 atom, 135 atom and 320 atom supercells. (b) The curved migration pathway followed by the diffusing O atom. (c) The magnetic moments of the nearest neighbor Co ions involved in the migration process.

The Co ions in the proximity of the migrating O ion are shown in Fig. 4(b). CoB remains coordinated to the migrating O ion (and the vacant site) through its transition from the initial to the final state. As such its magnetic moment remains low throughout. On the other hand, CoA and CoC are coordinated to the vacancy in the final and initial states respectively, which leads to a nearly symmetric change in μ678 and μ679 between the initial and final states. This change is not as abrupt as in other perovksites34,38 which is an indication of the covalent character of Co-O bonding.

OER with oxygen vacancies- While it has been inferred from theoretical calculations5,8 and experiments39,40 that SrCoO3 is one of the best materials for OER, a clear experimental verification remains difficult. Experimentally, it remains challenging to grow fully oxidized thin films of SrCoO3.12,32

ACS Paragon Plus Environment

10

Page 11 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Zhou et al.12 also noticed a reduced OER activity when phases other than the cubic phase were present. Therefore, it is important to understand how oxygen vacancies alter the surface properties of SrCoO3 and how they interact with OER intermediates. We consider an oxygen vacancy on the surface of the (001) plane and another subsurface vacant O site to see how this could alter the activity of the Co sites. For this part of the work we used the RPBE functional which is better for handling adsorption energies.4,5,26 The workfunction of the bare (001) surface is 6.00 eV (6.35 eV with PBEsol). This corresponds to 1.56 V vs SHE which indicates a strong oxidizing potential. When introducing an oxygen vacancy the workfunctions become 5.85 and 5.97 eV for the surface and subsurface sites, respectively.

Figure 5. The free energy for each reaction step on different Co sites for (a) surface and (b) subsurface VO. The surface and subsurface oxygen vacancies are shown in the left and right panels, respectively. The distinct active sites for the 2x2 surface in the presence of the vacancy are as indicated.

Fig. 5 shows the free energy profiles of the OER steps. When the electrode potential is 0 V all reaction steps are uphill. For surface VO, the Co sites shown in Fig. 5(a) exhibit different activities. While on both sites the step O∗ → HOO∗ is rate determining, we find that the energy required for this step increases at the site in the immediate vicinity of VO, leading to an overpotential of 0.7 V and 0.52 V for site1 and site2, respectively. This is 0.2~0.4 V higher than the perfect SrCoO3 surface. Similarly, subsurface VO

ACS Paragon Plus Environment

11

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 18

leads to a reduction in the activity of the Co site immediately above it (site1) for which the overpotential is 0.7 V, and rate determining step is the dissociation of a water molecule to HO∗ which is only slightly higher than the formation of HOO∗ . Moving away from the vacant site, the activity improves with overpotentials calculated to ~0.45 V.

The presence of vacancies causes a local distortion in the Co-O-Co bonds. For instance, for surface VO the Co-O-Co angle was found to be 164°. Additionally, as we showed earlier, VO donates its electrons to it neighboring Co ions, thus reducing them. These reduced surface Co sites along with the buckled Co-OCo bonds leads to a decrease in the overlap between the Co 3d and O 2p orbitals leading to a reduction in the electron transfer between the surface and the OER intermediates.40 This deterioration in activity due to the presence of oxygen vacancies have been recently suggested by Cheng et al.40

Figure 6. Surface VO formation energy and overpotential for OER plotted against strain. The formation energy increases with biaxial compressive strain while the overpotential remains close to the unstrained case.

Finding means to suppress the formation of these surface vacancies would be highly valuable. To this end, strain engineering has been successfully applied to tune electronic and thermodynamic properties of

ACS Paragon Plus Environment

12

Page 13 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

solids. In this case we find that the application of moderate biaxial compressive strain leads to an increase in surface VO formation energy (Fig. 6). Application of hydrostatic or compressive strain on perovskites straightens the B-O-B bonds.41,42 Compression which favors Co-O-Co angles closer to 180° competes with tendency of a vacancy to distort the lattice hence disfavoring its formation which increases to 1 eV at a -2% strain compared to 0.75 eV in the unstrained case. On the other hand, the OER activity in the moderately strained structures remains close to the ideal case with a predicted overpotential of 0.31 V. A similar approach was used recently to control oxygen vacancies in brownmillerite SrCoOx.43 While this increase in the formation energy does not completely eliminate surface oxygenated species, it should provide a way forward to achieve a higher capacity of the catalytic power of SrCoO3.

4. Conclusions In this work we have studied the electronic, thermodynamics and kinetic aspects of the oxygen vacancies in SrCoO3. We find that creation of an oxygen vacancy effectively donates two electrons to neighboring Co sites, thereby reducing them. These vacancies have low formation energies which are calculated to be 1.17-1.26 eV in the dilute limit. The migration barrier for a diffusing oxygen vacancy was found to be ~0.5 eV, which suggests that this defect is quite mobile relative to other typical perovskites.29,31,34 In accord with the high oxygen deficiencies reported in experiments, we find that in the context of divacancies this material has a tendency to form clustered VO which are aligned in a basal or apical configuration. Finally, we have investigated the effect of surface vacancies on the OER activity and we found oxygen vacancies are detrimental to the materials performance, in line with previous experiments.40 Applying biaxial compressive strain increases the energetic cost of forming these defects and does not negatively impact catalytic activity compared with the unstrained pristine perovskite (001) surface, suggesting a possible route to minimizing surface vacancies and thereby optimizing the catalytic performance.

ACS Paragon Plus Environment

13

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 18

ASSOCIATED CONTENT Supporting Information. Details of calculations and models used. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author * Sean C. Smith: [email protected]

ACKNOWLEDGEMENT This research was undertaken with the assistance of UNSW Australia SPF01 funding (SCS). We acknowledge generous allocations of supercomputing time at the Pawsey Supercomputing Centre via the Australian National Computational Merit Allocation Scheme (NCMAS project fr2) and the Energy and Resources Merit Allocation Scheme of the Pawsey Supercomputing Centre (project pawsey0111). Additional computational resources were provided by KAUST on the Shaheen II supercomputer.

REFERENCES (1)

Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. a.; Lewis, N. S. Chem. Rev. 2010, 110, 6446–6473.

(2)

Katsounaros, I.; Cherevko, S.; Zeradjanin, A. R.; Mayrhofer, K. J. J. Angew. Chemie Int. Ed. 2014, 53, 102–121.

(3)

Li, Y.-F.; Liu, Z.-P. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2016, 6, 47–64.

ACS Paragon Plus Environment

14

Page 15 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

(4)

Nørskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jónsson, H. J. Phys. Chem. B 2004, 108, 17886–17892.

(5)

Man, I. C.; Su, H.-Y.; Calle-Vallejo, F.; Hansen, H. a.; Martínez, J. I.; Inoglu, N. G.; Kitchin, J.; Jaramillo, T. F.; Nørskov, J. K.; Rossmeisl, J. ChemCatChem 2011, 3, 1159–1165.

(6)

Zhang, J.; Zhao, Z.; Xia, Z.; Dai, L. Nat. Nanotechnol. 2015, 10, 444–452.

(7)

Rossmeisl, J.; Logadottir, A.; Nørskov, J. K. Chem. Phys. 2005, 319, 178–184.

(8)

Calle-Vallejo, F.; Díaz-Morales, O. A.; Kolb, M. J.; Koper, M. T. M. ACS Catal. 2015, 5, 869– 873.

(9)

Muñoz, A.; de la Calle, C.; Alonso, J. A.; Botta, P. M.; Pardo, V.; Baldomir, D.; Rivas, J. Phys. Rev. B 2008, 78, 54404.

(10)

Suntivich, J.; May, K. J.; Gasteiger, H. A.; Goodenough, J. B.; Shao-Horn, Y. Science 2011, 334, 1383–1385.

(11)

Kuklja, M. M.; Mastrikov, Y. A.; Jansang, B.; Kotomin, E. a. Solid State Ionics 2013, 230, 21–26.

(12)

Zhou, W.; Zhao, M.; Liang, F.; Smith, S. C.; Zhu, Z. Mater. Horiz. 2015, 2, 495–501.

(13)

Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169–11186.

(14)

Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758–1775.

(15)

Perdew, J. P.; Ruzsinszky, A.; Csonka, G. I.; Vydrov, O. A.; Scuseria, G. E.; Constantin, L. A.; Zhou, X.; Burke, K. Phys. Rev. Lett. 2008, 100, 136406.

(16)

Hammer, B.; Hansen, L. B.; Nørskov, J. K. Phys. Rev. B 1999, 59, 7413–7421.

(17)

Xu, X.; Goddard, W. A. J. Phys. Chem. A 2004, 108, 2305–2313.

ACS Paragon Plus Environment

15

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(18)

Neugebauer, J.; Scheffler, M. Phys. Rev. B 1992, 46, 16067–16080.

(19)

Taguchi, H.; Shimada, M.; Koizumi, M. J. Solid State Chem. 1979, 29, 221–225.

(20)

Hoffmann, M.; Borisov, V. S.; Ostanin, S.; Mertig, I.; Hergert, W.; Ernst, A. Phys. Rev. B 2015,

Page 16 of 18

92, 94427. (21)

Mastrikov, Y. a; Kuklja, M. M.; Kotomin, E. a; Maier, J. Energy Environ. Sci. 2010, 3, 1544.

(22)

Bezdicka, P.; Wattiaux, A.; Grenier, J. C.; Pouchard, M.; Hagenmuller, P. Zeitschrift für Anorg. und Allg. Chemie 1993, 619, 7–12.

(23)

Long, Y.; Kaneko, Y.; Ishiwata, S.; Taguchi, Y.; Tokura, Y. J. Phys. Condens. Matter 2011, 23, 245601.

(24)

Momma, K.; Izumi, F. J. Appl. Crystallogr. 2011, 44, 1272–1276.

(25)

Wang, L.; Maxisch, T.; Ceder, G. Phys. Rev. B 2006, 73, 195107.

(26)

Lee, Y.-L.; Kleis, J.; Rossmeisl, J.; Morgan, D. Phys. Rev. B 2009, 80, 224101.

(27)

Chen, D.; Chen, C.; Zhang, Z.; Baiyee, Z. M.; Ciucci, F.; Shao, Z. ACS Appl. Mater. Interfaces 2015, 7, 8562–8571.

(28)

Castleton, C. W. M.; Mirbt, S. Phys. Rev. B 2004, 70, 195202.

(29)

Malashevich, A.; Ismail-Beigi, S. Phys. Rev. B 2015, 92, 144102.

(30)

Mastrikov, Y. A.; Merkle, R.; Heifets, E.; Kotomin, E. A.; Maier, J. J. Phys. Chem. C 2010, 114, 3017–3027.

(31)

Cuong, D. D.; Lee, B.; Choi, K. M.; Ahn, H.-S.; Han, S.; Lee, J. Phys. Rev. Lett. 2007, 98, 115503.

ACS Paragon Plus Environment

16

Page 17 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

(32)

Jeen, H.; Choi, W. S.; Freeland, J. W.; Ohta, H.; Jung, C. U.; Lee, H. N. Adv. Mater. 2013, 25, 3651–3656.

(33)

Jeen, H.; Bi, Z.; Choi, W. S.; Chisholm, M. F.; Bridges, C. a.; Paranthaman, M. P.; Lee, H. N. Adv. Mater. 2013, 25, 6459–6463.

(34)

Ritzmann, A. M.; Muñoz-García, A. B.; Pavone, M.; Keith, J. A.; Carter, E. A. Chem. Mater. 2013, 25, 3011–3019.

(35)

Jones, A.; Islam, M. S. J. Phys. Chem. C 2008, 112, 4455–4462.

(36)

Mitra, C.; Meyer, T.; Lee, H. N.; Reboredo, F. A. J. Chem. Phys. 2014, 141, 84710.

(37)

Kotomin, E. a.; Mastrikov, Y. a.; Kuklja, M. M.; Merkle, R.; Roytburd, a.; Maier, J. Solid State Ionics 2011, 188, 1–5.

(38)

Ritzmann, A. M.; Pavone, M.; Muñoz-García, A. B.; Keith, J. A.; Carter, E. A. J. Mater. Chem. A 2014, 2, 8060–8074.

(39)

Matsumoto, Y. J. Electrochem. Soc. 1980, 127, 811.

(40)

Cheng, X.; Fabbri, E.; Nachtegaal, M.; Castelli, I. E.; El Kazzi, M.; Haumont, R.; Marzari, N.; Schmidt, T. J. Chem. Mater. 2015, 27, 7662–7672.

(41)

Loa, I.; Adler, P.; Grzechnik, A.; Syassen, K.; Schwarz, U.; Hanfland, M.; Rozenberg, G. K.; Gorodetsky, P.; Pasternak, M. P. Phys. Rev. Lett. 2001, 87, 125501.

(42)

Ross, N. L.; Zhao, J.; Angel, R. J. J. Solid State Chem. 2004, 177, 3768–3775.

(43)

Petrie, J. R.; Mitra, C.; Jeen, H.; Choi, W. S.; Meyer, T. L.; Reboredo, F. A.; Freeland, J. W.; Eres, G.; Lee, H. N. Adv. Funct. Mater. 2016, 26, 1564–1570.

ACS Paragon Plus Environment

17

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 18

Table of Contents Graphic and Synopsis SrCoO3 is one of the best materials for oxygen evolution reaction. However, the presence of oxygen vacancies interferes with its catalytic properties leading to an increase in overpotentials from those predicted for the ideal surfaces. Compressive biaxial strain can suppress the formation of surface oxygen vacancies without a significant effect on the overpotential making this an attractive pathway to harness the material’s full catalytic power.

ACS Paragon Plus Environment

18